The overall aim of this project is to precisely determine the carcinogenicities of 4 related aflatoxins (AFB1, AFL, AFM1, AFL-M1) and to understand their relative carcinogenic potencies in terms of: patterns of cellular metabolism; maximum quantities and types of DNA adducts formed; persistence of individual adducts; and the relative frequencies with which each adduct forms and persists on specific oncogene nucleotide sites implicated in cell transformation and tumor initiation. The 4 aflatoxins form a model analogue series for examinig fundamental structure-activity relationships in cancer initiation by bulky carcinogens. Trout are perhaps the only developed vertebrate model with sufficient sensitivity to test nanogram doses of rare aflatoxins, and of sufficient low cost that such studies (which involve 7200-animal tumor experiments) are reasonably supportable. To achieve this goal there are several Specific Aims: 1) We wish to precisely determine the relative carcinogencities of 4 aflatoxins, and separately, of their 4 dihalide activated derivatives, by direct trout embryo mciroinjection. We will also determine the embryo distribution and maximum covalent DNA binding indices (CBI) for each aflatoxin, and use a statistical model to examine the hypothesis that differences in CBI alone predict relative tumorigenicities among these aflatoxins or, separately, their dihalides. 2) We also examine the relationship between relative carcinogenicites among aflatoxins, and the exposure protocols (embryo vs. dietary) used to determine the relative carcinogencities. 3) Mechanism studies will examine the extent to which relative carcinogenicities reflect individual adduct structures, their persistence in vivo and in vitro, and the relative frequencies with which each aflatoxin adduct is formed at various guanyl sites in a cloned c-ras protooncogene including codon 12 guanines. Experiments will be conducted to compare site-specific rates of guanyl depurination among aflatoxin adducts, and to determine the relative potencies for transformation of NIH 3T3 cells among c-ras plasmids containing known levels of adducts, or apurinic sites, at each guanine including condon 12 sites. These studies should considerably improve our understanding of the biochemical and molecular mechanisms underlying differential carcinogenicities and mutagencities among related bulky carcinogens. They may also provide unique information on a fundamental question in carcinogenesis, namely whether c-ras activation is causative for tumor initiation.